Why We Built Our Own EDC

Written By Caleb Ruth

Series: Choosing your EDC, Part 1

The Research Allies Origin Story

It was the late 1990s. Fluorescent lights flickered. Flies buzzed. In the ICU of an under-resourced teaching hospital in Ljubljana, Slovenia, I watched a neurosurgery resident type patient data into a new IBM ThinkPad while the attending physicians huddled around a clipboard. The man on the cot in front of us had a hole drilled in his skull. A catheter ran from his bandaged head into an open glass jar on the floor.

I wasn't there as a physician. I was there to make sure what happened in that room would count — that the data would survive to help future patients.

The system was called the Traumatic Brain Injury Survey — TBIS — and it was one of the first electronic data capture (EDC) tools deployed in post-Soviet Eastern Europe. Our goal was to train physicians on evidence-based treatment protocols and collect data that could validate and refine those protocols for future patients. We built it in the age of dial-up Internet, when research data was almost exclusively captured on paper and Netflix was still a DVD-by-mail service.

TBIS worked. It ran in a dozen teaching hospitals from Estonia to Croatia for several years and spawned three separate data tools. More importantly, it showed me something I've never forgotten: a well-built data system, deployed in a resource-limited setting, can improve the science behind the care.

That belief is why Research Allies exists.

The Gap That Wouldn't Go Away

Fast-forward to 2015. I got a call from the Mehta Research Group at Cornell University. They were preparing to run clinical trials in rural India and Ecuador — nutritional intervention studies with hundreds of participants, complex eligibility criteria, multilingual field teams, and almost no reliable internet access.

They had already looked at the available options. REDCap, the most widely used EDC platform in academic research, hadn't yet released its mobile app when their first projects launched in 2013. Even later, REDCap's architecture was fundamentally web-based — when connectivity dropped, so did the system. ODK (Open Data Kit) offered offline capability, but it had security gaps that made it unsuitable for regulated clinical research. Custom tools were expensive and couldn't be reused across studies.

What the Cornell team needed didn't exist: a mobile-first EDC that could function completely offline, enforce complex protocol requirements automatically, and be secure enough for remote field work and HIPAA compliance.

So they asked me to build it.

What We Actually Needed

The problem with most EDC tools isn't that they're bad at what they do. It's that they're built for a specific context — usually a well-funded institution with reliable infrastructure — and then adapted, sometimes poorly, for everything else.

What we needed was different in three ways.

First, offline-first by design. Field teams in rural India weren't collecting data in a hospital with fiber Internet. They were working in homes and community clinics where connectivity was unpredictable and sometimes nonexistent. The system had to work with zero network access and synchronize later when a connection became available. That's not a feature you bolt onto a web-based architecture. It has to be designed in from the start.

Second, protocol enforcement — not just data collection. A field research assistant shouldn't have to remember which forms to complete at a midline visit, or manually verify whether a participant is eligible to enroll, or track which consent forms still need signatures. The system should know the protocol and enforce it automatically. That means the EDC isn't just a data entry interface — it's the study running itself.

Third, compliance without complexity. Clinical research data is subject to strict regulatory requirements: audit trails, electronic signatures, data security, access controls. These can't be afterthoughts. But they also can't require a team of IT specialists to maintain in the field.

What We Built

We called it ConnEDCt.

The name was intentional: EDC for connecting data in disconnected communities. ConnEDCt was built on FileMaker, a cross-platform relational database platform from Claris International, which gave us two things we needed immediately — rapid development and reliable deployment across iPads and laptops without maintaining separate codebases.

The result was a system that ran natively on iPads in the field and synchronized data to a cloud server whenever connectivity allowed. Study protocol configurations — which forms to present at which visits, which eligibility criteria to evaluate, how to assign randomization — were defined by study managers and loaded into the system, minimizing the programming required for each new study.

We deployed ConnEDCt across five research investigations over the next several years:

  • A clinical study tracking fever patients in Ecuador, where multiple team members on separate devices needed to sync data continuously during patient visits using mobile SIM cards

  • A randomized controlled trial of biofortified pearl millet in Mumbai, where over 400 children were enrolled, randomized, and tracked across nine monthly follow-up visits entirely in Hindi

  • A periconceptional surveillance program in Southern India monitoring more than 2,400 households and 2,876 women, with a hierarchical participant ID system designed to handle complex household structures

  • A second RCT in South India tracking mother-infant dyads across a 9-month follow-up with two-level randomization

  • A cohort study of KSHV in Uganda

By the time we published the results in the Journal of Medical Internet Research in 2020, ConnEDCt had supported more than 3,000 participants, 13,000 participant encounters, and 55,000 completed case report forms — all in settings where most EDC tools wouldn't function at all.

What It Became

ConnEDCt was a research project. Research Allies is the company that turned what we learned into a product.

The same architecture — mobile-first, offline-capable, protocol-enforcing, compliance-ready — is now the foundation of a purpose-built platform for investigators who are running the kinds of studies that mainstream EDC tools weren't designed for. Complex protocols. Multilingual teams. Remote environments. Limited institutional support.

We didn't build Research Allies because we thought we could make a better REDCap. We built it because we kept running into investigators who needed something REDCap couldn't give them. We already knew what that something was, because we'd spent years building and running it ourselves.

That's the origin story.

In the next article in this series, we'll step back and answer a more basic question: what actually is an EDC, and do you even need one?

Caleb Ruth is the co-founder of Research Allies and the original developer of ConnEDCt. The problem he set out to solve in post-Soviet Eastern Europe in the late 1990s is the same one he's still solving today.

Sources

  • Ruth, C.J. ConnEDCt, A mobile-first framework for clinical electronic data capture [Master's thesis, Boston University] (2021). OpenBU. https://open.bu.edu/handle/2144/42353

  • Ruth, C.J., Huey, S, et al., An Electronic Data Capture Framework (ConnEDCt) for Global and Public Health Research: Design and Implementation. Journal of Medical Internet Research, 2020. 22(8)

Caleb Ruth